the fact, that it in all cases effervesces slightly when dissolved in mu riatic acid.

We may sum up, therefore, the composition of this true basic carbonate of palladium as follows:

:

The formula deducible from these results is 10. Pd O + CO2

+10. HO, which gives the following numerical result :

It is not impossible but this body may be in reality a mixture of a less basic carbonate, with the true hydrated oxide, but I consider that the remarkable constancy of composition, indicated by so many specimens, prepared at different times, giving results so closely coinciding, argues very strongly in favour of its definite nature. It occured to me also, that the content in carbonic acid might arise from the presence of carbonate of soda, but I satisfied myself that although it is very difficult to obtain specimens which do not after ignition yield traces of alkali, yet it is never present in such quantity as could give the above results, when the freshly precipitated substance has been properly washed.

The properties of the suboxide of palladium, now first definitely found, are not very distinct. The existence of this suboxide had been long suspected, especially from the fact, that by heating to dull redness in contact with air, metallic palladium becomes coloured blue or green on the surface, which colours are removed by violent ignition. Berzelius found, however, that this colouring was not attended with any increase of weight, which arises, however, from the minute quantity of oxide formed, the colours being those of thin plates; but there can be now no doubt but that formation of suboxide does so occur. That the black powder which is left by the dull ignition of the basic carbonate is really a definite compound, is strongly supported by the fact of the accuracy with which the decomposition stops at its formation, and by the analogy of the subchloride, to be hereafter described, it gains additional force. Yet, I have not been able to combine in any way this suboxide with acids. By contact with them, it gives an ordinary salt of the protoxide and

R

metal.

It is possible, however, that in future trials I may be more successful.

It is known that by the addition of a caustic fixed alkali to a salt of palladium a precipitate is obtained, which redissolves in a great excess of the alkali. The precipitate is in this case a basic salt, not the hydrated oxide, and with difficulty contains traces of the alkali, from which it is with difficulty freed by washing. The soluble alkaline compound cannot be obtained in a definite form by any process that I have as yet tried. By evaporation to dryness, the oxide separates anhydrous, and retaining a small quantity of alkali.

When the basic carbonate of palladium (hydrated oxide) is diffused through water of ammonia, it partly dissolves, giving a powder and a yellow-coloured liquor, which when evaporated, dries down to a bright yellow diliquescent mass. When heated, this decomposes with slight deflagration and copious disengagement of gas, leaving metallic palladium. The brown powder also contains ammonia, and when heated gives it off with water, and the metal is reduced. I can, however, at present only indicate the existence of these two bodies, as the analytical results which I have obtained respecting them are too discordant to admit of my assigning any definite formula for their constitution. The soluble compound I conceive to arise from the ammonia acting on the carbonate, as it does on any other palladium salt, and the insoluble to be the product of the ammonia on the oxide which is present in excess. I reserve these bodies, therefore, as objects of future study.

Chlorides of Palladium.

The properties of the ordinary protochloride of palladium have been for the most part so fully described by those chemists that have occupied themselves with the study of this metal, that I shall notice it but briefly. From a strong solution, it crystallizes in prismatic needles which are very deliquescent. These crystals were

found to contain two atoms of water of crystallization, which they lose by a gentle heat.

The action of a high temperature on protochloride of palladium developes some facts of considerable interest. It is not so reduced to the metallic state unless by very violent ignition, but just at a red heat it melts and begins to evolve chlorine, which continues until it has parted with one-half of that which it contains. The liquified mass which remains is a true subchloride, which is not further acted upon unless the heat be very much increased.

The following numerical results will render this decomposition evident :

A. 29 881 grains of the crystallized protochloride being carefully dried, as long as they gave off any traces of watery vapour, were found to have lost 5.247 grains, 17.56 per cent.

The dried mass was heated until it had completely fused. It was then dull red, In this state it was found to weigh 22.055 grains, having given off 2:577 grains of chlorine, or 8.63 per cent.

This was next kept melted at a bright red heat, until it appeared to cease giving off any gas. It then weighed 19-632 grains, having lost in addition 2-423 grains, or 8.11 per cent. of chlorine.

This residue was now fully ignited with some carbonate of ammonia, until the metallic palladium remaining appeared to be quite pure; this then weighed 14.554 grains, or 48.71 per cent., the quantity of chlorine abandoned in this final stage having been 5.078 grains, or 16.99 per cent.

The quantity of palladium, and the total quantity of chlorine and water, show that the salt in its crystalline condition has the formula Pd Cl + 2 H O, by which we have

The relation between the proportions of chlorine which were evolved at the different periods, were as 8'11, 8'63, and 16.99. I do not attach much importance to the two first being so nearly equal, but to the fact that the quantity which was expelled by the heat, was sensibly equal to half the total quantity in the salt.

B. 58-919 grains of dried protochloride were heated in a porcelain crucible to full redness, until it fused without any disengagement of gas. The residual subchloride weighed 47.782 grains, or

81.13 per cent.

C. 138-397 grains were fused in a porcelain crucible, and kept at a full red heat until all effervescence from loss of chlorine ceased. When cold it weighed 110.185 grains, equivalent to 80 41 per cent. These results fully prove that the loss of chlorine which the protochloride undergoes when kept for some time fused at a full red heat, is perfectly definite; and also that the loss represents one half of the chloride which the salt contains. Thus,

or 80.05 per cent. The substance formed is a true subchloride analogous to calomel, or to subchloride of copper, and its formula is Pd2 Cl.

Its

The fused mass obtained by the methods now described, is of a deep red brown colour, and highly chrystalline in structure. powder is light red. It deliquesces rapidly, and becomes dark coloured from the separation of metallic palladium and the formation of protochloride. This change is effected almost instantly by contact with water, or solutions of sal-ammoniac, or iodide of potassium, also

by water of ammonia. It is, however, not all decomposed; the quantity of metallic palladium which separates, I have found to be but from one-fifth to one-sixth of that which the subchloride contained. The liquor formed contains, therefore, both the subchloride and the protochloride dissolved together. The action of reagents on this liquor, however, does not differ materially from that produced with solutions of the protochloride. The liquor is much darker coloured than a solution of protochloride of the same strength should be, and is rendered turbid by dilution with more water. The first action of ammonia appears to be, the formation of a white compound, which is, however, almost instantly broken up into a pinkish ammonia-protochloride, and metallic palladium.

In the double salts formed by the protochloride of palladium with the chlorides of the alkaline metals, I have found the similarity of constitution, so usual between the compounds of ammonium and potassium, to be violated. The double chloride of palladium and potassium has been fully shewn by Berzelius to have the formula Pd. Cl + K. Cl, and not to contain any water of crystallization; but the double chloride of palladium and ammonium retains an atom of water when crystallized. I examined a quantity of this salt which had formed long rectangular prisms of an olive colour with a rich bronze lustre. They were quite free from any foreign impurities. When heated, they yielded water, muriatic acid, sal-ammonia, and left metallic palladium.

By a very cautious application of heat, the water may be completely expelled. From many experiments its quantity was found to be from 5.52 to 5'95 per cent., and the residual palladium was ascertained to be from 35.56 to 35.27 in 100 of the crystals. These numbers indicate the formula Pd Cl + N H1. Cl + H O, which gives

If the salt were anhydrous, it should yield 37.51 per cent of metal. It is only in consequence of its differing from the potassium salt that I deem this body worthy of notice here.

Of the Oxychloride of Palladium.

When a solution of chloride of palladium is partially precipitated by means of a solution of potash or of soda, care being taken that the metallic chloride shall still be present in considerable excess, a dark brown powder is obtained, which is a definite oxychloride of palladium.

When dried in a stove at a temperature of 150°, its properties are as follows if it be heated, it gives off water, and at a full red heat chlorine and oxygen, leaving behind a mixture of subchloride, suboxide, and metal. It dissolves in dilute acids, giving a mixture of protochloride and an ordinary palladium salt of the acid used. Its analysis was conducted as follows:

A. 40 639 grains fused with carbonate of soda and the saline mass dissolved in water, left 28.508 grains of palladium, purely metallic, being equal to 69.80 per cent.

The solution acidulated by nitric acid, and precipitated by nitrate of silver, gave 18·103 grains of chloride of silver, or 44.55 per cent., containing 10.99 of chlorine.

B. 67.543 grains of another specimen were heated over a spirit lamp as long as any watery vapour came off, but not so high as to expel any traces of chlorine or oxygen. The dry mass which remained weighed 59,977 grains being 88.62 per cent. The loss of water had thus been 11.38 per cent.

This dry residue was then vividly ignited, and a lump of carbonate of ammonia introduced to favour the seperation of the chlorine. The metallic palladium which remained weighed 47-442 grains, or 70.25 per cent.

These results lead to the formula Pd Cl + 3 Pd O + 4 H O, which gives,

It is therefore quite analogous to the ordinary oxychloride of copper.

Of the Ammonia-chlorides of Palladium.

It is well known, that on adding water of ammonia to a solution of chloride of palladium, a pink-coloured precipitate is produced, which by boiling dissolves, giving a brownish yellow liquor, from which, on cooling, a crystalline yellow substance separates. These two bodies have the same per cent, composition, expressed by the formula Pd Cl + NH3. Of this I need not detail any evidence, as it has been fully established by the labours of Berzelius, and quite recently by the experiments of Fehling.

By means of an excess of ammonia, the pink red precipitate which first forms may be totally redissolved, giving a colourless solution, from which by evaporation, a salt is deposited on cooling, in colourless rectangular prisms. The existence of this salt has been long known, and as its analysis has been recently published by Fehling, I need not detail any of my own experimental results